Re: production use of reiserfs

Re: production use of reiserfs

[Hans Reiser]

Nikita Danilov wrote:

>ivan writes:
> > Hi ,
> > Thanks for the answer!.
> > 
> > It is the info.
> > I hope it is all that I have.
> > And it is from the last crashed system.
> > 
> > /etc/fstat:
> > 
> > /dev/hda2               /                       reiserfs defaults        1 1
> > /dev/hda1               /boot                   reiserfs defaults        1 2
> > none                    /dev/pts                devpts  gid=5,mode=620  0 0
> > none                    /proc                   proc    defaults        0 0
> > none                    /dev/shm                tmpfs   defaults        0 0
> > /dev/hda3               swap                    swap    defaults        0 0
> > /dev/hdc2               /mnt/diske              reiserfs defaults        1 2
> > 
> > 
> > 
> > [root@tp1 root]# dmesg
> > Linux version 2.4.20-8 (bhcompile@porky.devel.redhat.com) (gcc version 3.2.2 
> > 20030222 (Red Hat Linux 3.2.2-5)) #1 Thu Mar 13 17:54:28 EST 2003
> > BIOS-provided physical RAM map:
>
>We need kernel messages issued during crash. It is possible (albeit not
>necessary) that they managed to get into /var/log/messages, if crashes
>are severe enough you will have to compile serial console support into
>kernel, attach null-modem to the serial port of the target machine and
>run something similar to minicom(1) on the other end to collect kernel
>messages during crash.
>
> >  BIOS-e820: 0000000000000000 - 00000000000a0000 (usable)
> >  BIOS-e820: 00000000000f0000 - 0000000000100000 (reserved)
>
>[...]
>
> > >  > Also it is possible to use redhat AS 3.0, but we do not find any info
> > >  > how to instsall it with reiserfs. Exist any problems on this distro?
> > >
> > > It is much easier for us to track and deal with problems in the stock
> > > kernel, but look at the http://www.namesys.com/support.html :)
> > 
> > I readet it, but it is for problems.
> > I can not figure exactly my problem.
> > It is not one time crash.
>
>Yes, I meant money can buy you better support. :)
>
> > 
> > I am searching info to know if reiserfs is production ready.
>
>People are using it successfully. It is standard file system in some
>distributions. It has been used on some very large highly loaded
>servers.
>
> > And also to know if exists any problems in redhat implementation for reiserfs.
>
>That's hard to tell, because we don't know whether redhat modified
>reiserfs code in any way. What you can do is to download 2.4.20 kernel
>from kernel.org and do "diff -rup" between /usr/src/linux/fs/reiserfs
>and /stock-2.4.20-kernel/fs/reiserfs.
>  
>
this is not likely to show any differences, sources of bugs are likely 
to be outside reiserfs.

I can tell you that I don't particularly trust reiserfs on redhat 
kernels because they don't apply reiserfs bug fix patches and generally 
don't mind breaking reiserfs.  Please download  the latest official 
kernel if you want to be sure to have all bugfixes we have made.  The 
latest official kernel is very solid for reiserfs, as is the suse kernel 
(excepting the extended attributes patch SuSE wrote which you can 
disable in the kernel configuration).  This particular kernel you have 
might be reasonably stable because 2.4.20 was reasonably stable for us, 
but as I said the latest official kernel generates no bug reports on 
filesystems created from scratch on it.   We would investigate and fix 
your bug report for free except that RedHat will not apply it if we 
generate a patch, so we would only do it for a fee.

> > 
> > >
>
>Nikita.
>
> > 
> > regards,
> > ivan.
>
>
>  
>


-- 
Hans

production use of reiserfs

[ivan]

Hi all,

We are collecting our statistic for the last 9 m. of using linux + postgreSQL 
as oracle+windows alternative in production.

We are using linux redhat (7.3 and 9.0) + reiserfs (it was recomendet as 
stable and fast).

For this time we have 6 filesystem crashes (all after powerdown) with total 
database lose.

It is very bad situation because for this period we do not have any problems 
with windows + oracle.

So it is now time to see if resierfs is production ready or we need to change 
the filesystem (or OS distro).

My question:

exist somewath special in reiserfs setup on redhat, that we need to change?
We are using the standart install (linux reiserfs).

We do not make changes on kernel or setup.

Is it good idea to change redhat?

Also it is possible to use redhat AS 3.0, but we do not find any info how to 
instsall it with reiserfs. Exist any problems on this distro?

regards,
ivan.

Re: production use of reiserfs

[Nikita Danilov]

ivan writes:
 > Hi all,

Hello,

 > 
 > We are collecting our statistic for the last 9 m. of using linux + postgreSQL 
 > as oracle+windows alternative in production.
 > 
 > We are using linux redhat (7.3 and 9.0) + reiserfs (it was recomendet as 
 > stable and fast).
 > 
 > For this time we have 6 filesystem crashes (all after powerdown) with total 
 > database lose.
 > 
 > It is very bad situation because for this period we do not have any problems 
 > with windows + oracle.

Please provide more info: kernel messages, mount options, kernel version
(cat /proc/version), reiserfs-related portion of kernel .config, etc.

 > 
 > So it is now time to see if resierfs is production ready or we need to change 
 > the filesystem (or OS distro).
 > 
 > My question:
 > 
 > exist somewath special in reiserfs setup on redhat, that we need to change?
 > We are using the standart install (linux reiserfs).
 > 
 > We do not make changes on kernel or setup.
 > 
 > Is it good idea to change redhat?
 > 
 > Also it is possible to use redhat AS 3.0, but we do not find any info how to 
 > instsall it with reiserfs. Exist any problems on this distro?

It is much easier for us to track and deal with problems in the stock
kernel, but look at the http://www.namesys.com/support.html :)

 > 
 > regards,
 > ivan.

Nikita.

Re: production use of reiserfs

[ivan]

Hi ,
Thanks for the answer!.

It is the info.
I hope it is all that I have.
And it is from the last crashed system.

/etc/fstat:

/dev/hda2               /                       reiserfs defaults        1 1
/dev/hda1               /boot                   reiserfs defaults        1 2
none                    /dev/pts                devpts  gid=5,mode=620  0 0
none                    /proc                   proc    defaults        0 0
none                    /dev/shm                tmpfs   defaults        0 0
/dev/hda3               swap                    swap    defaults        0 0
/dev/hdc2               /mnt/diske              reiserfs defaults        1 2



[root@tp1 root]# dmesg
Linux version 2.4.20-8 (bhcompile@porky.devel.redhat.com) (gcc version 3.2.2 
20030222 (Red Hat Linux 3.2.2-5)) #1 Thu Mar 13 17:54:28 EST 2003
BIOS-provided physical RAM map:
 BIOS-e820: 0000000000000000 - 00000000000a0000 (usable)
 BIOS-e820: 00000000000f0000 - 0000000000100000 (reserved)
 BIOS-e820: 0000000000100000 - 000000001ffeb000 (usable)
 BIOS-e820: 000000001ffeb000 - 000000001ffef000 (ACPI data)
 BIOS-e820: 000000001ffef000 - 000000001ffff000 (reserved)
 BIOS-e820: 000000001ffff000 - 0000000020000000 (ACPI NVS)
 BIOS-e820: 00000000fff80000 - 0000000100000000 (reserved)
0MB HIGHMEM available.
511MB LOWMEM available.
On node 0 totalpages: 131051
zone(0): 4096 pages.
zone(1): 126955 pages.
zone(2): 0 pages.
Kernel command line: ro root=/dev/hda2
Initializing CPU#0
Detected 938.034 MHz processor.
Console: colour VGA+ 80x25
Calibrating delay loop... 1874.32 BogoMIPS
Memory: 511304k/524204k available (1347k kernel code, 10340k reserved, 999k 
data, 132k init, 0k highmem)
Dentry cache hash table entries: 65536 (order: 7, 524288 bytes)
Inode cache hash table entries: 32768 (order: 6, 262144 bytes)
Mount cache hash table entries: 512 (order: 0, 4096 bytes)
Buffer-cache hash table entries: 32768 (order: 5, 131072 bytes)
Page-cache hash table entries: 131072 (order: 7, 524288 bytes)
CPU: L1 I cache: 16K, L1 D cache: 16K
CPU: L2 cache: 256K
Intel machine check architecture supported.
Intel machine check reporting enabled on CPU#0.
CPU:     After generic, caps: 0383f9ff 00000000 00000000 00000000
CPU:             Common caps: 0383f9ff 00000000 00000000 00000000
CPU: Intel Pentium III (Coppermine) stepping 06
Enabling fast FPU save and restore... done.
Enabling unmasked SIMD FPU exception support... done.
Checking 'hlt' instruction... OK.
POSIX conformance testing by UNIFIX
mtrr: v1.40 (20010327) Richard Gooch (rgooch@atnf.csiro.au)
mtrr: detected mtrr type: Intel
PCI: PCI BIOS revision 2.10 entry at 0xf0df0, last bus=2
PCI: Using configuration type 1
PCI: Probing PCI hardware
Transparent bridge - Intel Corp. 82801BA/CA/DB PCI Bridge
PCI: Using IRQ router PIIX [8086/2440] at 00:1f.0
isapnp: Scanning for PnP cards...
isapnp: No Plug & Play device found
Linux NET4.0 for Linux 2.4
Based upon Swansea University Computer Society NET3.039
Initializing RT netlink socket
apm: BIOS version 1.2 Flags 0x03 (Driver version 1.16)
Starting kswapd
VFS: Disk quotas vdquot_6.5.1
Detected PS/2 Mouse Port.
pty: 2048 Unix98 ptys configured
Serial driver version 5.05c (2001-07-08) with MANY_PORTS MULTIPORT SHARE_IRQ 
SERIAL_PCI ISAPNP enabled
Real Time Clock Driver v1.10e
Floppy drive(s): fd0 is 1.44M
FDC 0 is a post-1991 82077
NET4: Frame Diverter 0.46
RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize
Uniform Multi-Platform E-IDE driver Revision: 7.00beta-2.4
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
ICH2: IDE controller at PCI slot 00:1f.1
ICH2: chipset revision 2
ICH2: not 100% native mode: will probe irqs later
    ide0: BM-DMA at 0xb800-0xb807, BIOS settings: hda:DMA, hdb:pio
    ide1: BM-DMA at 0xb808-0xb80f, BIOS settings: hdc:DMA, hdd:pio
hda: MAXTOR 6L040J2, ATA DISK drive
blk: queue c03c9f40, I/O limit 4095Mb (mask 0xffffffff)
hdc: Maxtor 2B020H1, ATA DISK drive
blk: queue c03ca3a0, I/O limit 4095Mb (mask 0xffffffff)
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
ide1 at 0x170-0x177,0x376 on irq 15
hda: host protected area => 1
hda: 78177792 sectors (40027 MB) w/1818KiB Cache, CHS=4866/255/63, UDMA(33)
hdc: host protected area => 1
hdc: 40020624 sectors (20491 MB) w/2048KiB Cache, CHS=39703/16/63, UDMA(100)
ide-floppy driver 0.99.newide
Partition check:
 hda: hda1 hda2 hda3
 hdc: [PTBL] [2491/255/63] hdc1 hdc2 hdc3
ide-floppy driver 0.99.newide
md: md driver 0.90.0 MAX_MD_DEVS=256, MD_SB_DISKS=27
md: Autodetecting RAID arrays.
md: autorun ...
md: ... autorun DONE.
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP, IGMP
IP: routing cache hash table of 4096 buckets, 32Kbytes
TCP: Hash tables configured (established 32768 bind 65536)
Linux IP multicast router 0.06 plus PIM-SM
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
RAMDISK: Compressed image found at block 0
Freeing initrd memory: 179k freed
VFS: Mounted root (ext2 filesystem).
reiserfs: checking transaction log (device 03:02) ...
Using r5 hash to sort names
ReiserFS version 3.6.25
Freeing unused kernel memory: 132k freed
Adding Swap: 1020116k swap-space (priority -1)
reiserfs: checking transaction log (device 03:01) ...
Using r5 hash to sort names
ReiserFS version 3.6.25
reiserfs: checking transaction log (device 16:02) ...
Using r5 hash to sort names
ReiserFS version 3.6.25
ip_tables: (C) 2000-2002 Netfilter core team
8139too Fast Ethernet driver 0.9.26
PCI: Found IRQ 4 for device 02:0d.0
divert: allocating divert_blk for eth0
eth0: RealTek RTL8139 Fast Ethernet at 0xe0ada000, 00:c0:26:25:7b:30, IRQ 4
eth0:  Identified 8139 chip type 'RTL-8139C'
eth0: Setting 100mbps full-duplex based on auto-negotiated partner ability 
45e1.
lp: driver loaded but no devices found


[root@tp1 root]# cat /proc/version
Linux version 2.4.20-8 (bhcompile@porky.devel.redhat.com) (gcc version 3.2.2 
20030222 (Red Hat Linux 3.2.2-5)) #1 Thu Mar 13 17:54:28 EST 2003



On Friday 28 November 2003 11:48 am, Nikita Danilov wrote:
> ivan writes:
>  > Hi all,
>
> Hello,
>
>  > We are collecting our statistic for the last 9 m. of using linux +
>  > postgreSQL as oracle+windows alternative in production.
>  >
>  > We are using linux redhat (7.3 and 9.0) + reiserfs (it was recomendet as
>  > stable and fast).
>  >
>  > For this time we have 6 filesystem crashes (all after powerdown) with
>  > total database lose.
>  >
>  > It is very bad situation because for this period we do not have any
>  > problems with windows + oracle.
>
> Please provide more info: kernel messages, mount options, kernel version
> (cat /proc/version), reiserfs-related portion of kernel .config, etc.
>
>  > So it is now time to see if resierfs is production ready or we need to
>  > change the filesystem (or OS distro).
>  >
>  > My question:
>  >
>  > exist somewath special in reiserfs setup on redhat, that we need to
>  > change? We are using the standart install (linux reiserfs).
>  >
>  > We do not make changes on kernel or setup.
>  >
>  > Is it good idea to change redhat?
>  >
>  > Also it is possible to use redhat AS 3.0, but we do not find any info
>  > how to instsall it with reiserfs. Exist any problems on this distro?
>
> It is much easier for us to track and deal with problems in the stock
> kernel, but look at the http://www.namesys.com/support.html :)

I readet it, but it is for problems.
I can not figure exactly my problem.
It is not one time crash.

I am searching info to know if reiserfs is production ready.
And also to know if exists any problems in redhat implementation for reiserfs.

>
>  > regards,
>  > ivan.
>
> Nikita.

regards,
ivan.

Re: production use of reiserfs

[Nikita Danilov]

ivan writes:
 > Hi ,
 > Thanks for the answer!.
 > 
 > It is the info.
 > I hope it is all that I have.
 > And it is from the last crashed system.
 > 
 > /etc/fstat:
 > 
 > /dev/hda2               /                       reiserfs defaults        1 1
 > /dev/hda1               /boot                   reiserfs defaults        1 2
 > none                    /dev/pts                devpts  gid=5,mode=620  0 0
 > none                    /proc                   proc    defaults        0 0
 > none                    /dev/shm                tmpfs   defaults        0 0
 > /dev/hda3               swap                    swap    defaults        0 0
 > /dev/hdc2               /mnt/diske              reiserfs defaults        1 2
 > 
 > 
 > 
 > [root@tp1 root]# dmesg
 > Linux version 2.4.20-8 (bhcompile@porky.devel.redhat.com) (gcc version 3.2.2 
 > 20030222 (Red Hat Linux 3.2.2-5)) #1 Thu Mar 13 17:54:28 EST 2003
 > BIOS-provided physical RAM map:

We need kernel messages issued during crash. It is possible (albeit not
necessary) that they managed to get into /var/log/messages, if crashes
are severe enough you will have to compile serial console support into
kernel, attach null-modem to the serial port of the target machine and
run something similar to minicom(1) on the other end to collect kernel
messages during crash.

 >  BIOS-e820: 0000000000000000 - 00000000000a0000 (usable)
 >  BIOS-e820: 00000000000f0000 - 0000000000100000 (reserved)

[...]

 > >  > Also it is possible to use redhat AS 3.0, but we do not find any info
 > >  > how to instsall it with reiserfs. Exist any problems on this distro?
 > >
 > > It is much easier for us to track and deal with problems in the stock
 > > kernel, but look at the http://www.namesys.com/support.html :)
 > 
 > I readet it, but it is for problems.
 > I can not figure exactly my problem.
 > It is not one time crash.

Yes, I meant money can buy you better support. :)

 > 
 > I am searching info to know if reiserfs is production ready.

People are using it successfully. It is standard file system in some
distributions. It has been used on some very large highly loaded
servers.

 > And also to know if exists any problems in redhat implementation for reiserfs.

That's hard to tell, because we don't know whether redhat modified
reiserfs code in any way. What you can do is to download 2.4.20 kernel
from kernel.org and do "diff -rup" between /usr/src/linux/fs/reiserfs
and /stock-2.4.20-kernel/fs/reiserfs.

 > 
 > >

Nikita.

 > 
 > regards,
 > ivan.

Re: production use of reiserfs

[Dieter Nützel]

Am Freitag, 28. November 2003 12:49 schrieb Nikita Danilov:
> ivan writes:

> Yes, I meant money can buy you better support. :)
>
>  > I am searching info to know if reiserfs is production ready.
>
> People are using it successfully. It is standard file system in some
> distributions. It has been used on some very large highly loaded
> servers.
>
>  > And also to know if exists any problems in redhat implementation for
>  > reiserfs.
>
> That's hard to tell, because we don't know whether redhat modified
> reiserfs code in any way. What you can do is to download 2.4.20 kernel
> from kernel.org and do "diff -rup" between /usr/src/linux/fs/reiserfs
> and /stock-2.4.20-kernel/fs/reiserfs.

Maybe RedHat Linux whipping the file systems with to OLD reiserfsprogs after 
crash?

Which version (reiserfsck -V) do you have?

Greetings,
	Dieter

Re: production use of reiserfs

[Chris Mason]

On Fri, 2003-11-28 at 00:22, Hans Reiser wrote:
> >
> this is not likely to show any differences, sources of bugs are likely 
> to be outside reiserfs.
> 
> I can tell you that I don't particularly trust reiserfs on redhat 
> kernels because they don't apply reiserfs bug fix patches and generally 
> don't mind breaking reiserfs.  Please download  the latest official 
> kernel if you want to be sure to have all bugfixes we have made.  The 
> latest official kernel is very solid for reiserfs, as is the suse kernel 
> (excepting the extended attributes patch SuSE wrote which you can 
> disable in the kernel configuration).  This particular kernel you have 
> might be reasonably stable because 2.4.20 was reasonably stable for us, 
> but as I said the latest official kernel generates no bug reports on 
> filesystems created from scratch on it.   We would investigate and fix 
> your bug report for free except that RedHat will not apply it if we 
> generate a patch, so we would only do it for a fee.

Hans, please send along details for the xattr bugs you think are in the
suse kernel.  We have no open bug reports for them, and they work pretty
well.

-chris

Re: production use of reiserfs

[Redeeman]

this is strange, i feel that the linux filesystems is much more unstable
after a power failure than windows filesystens, but in advance they are
much faster. but ofcourse, windows filesystems are made to be crash
proof.

i have run into a crash with reiserfs, it didnt work, rebuild tree fixed
it perfect though

On Fri, 2003-11-28 at 07:46, ivan wrote:
> Hi all,
> 
> We are collecting our statistic for the last 9 m. of using linux + postgreSQL 
> as oracle+windows alternative in production.
> 
> We are using linux redhat (7.3 and 9.0) + reiserfs (it was recomendet as 
> stable and fast).
> 
> For this time we have 6 filesystem crashes (all after powerdown) with total 
> database lose.
> 
> It is very bad situation because for this period we do not have any problems 
> with windows + oracle.
> 
> So it is now time to see if resierfs is production ready or we need to change 
> the filesystem (or OS distro).
> 
> My question:
> 
> exist somewath special in reiserfs setup on redhat, that we need to change?
> We are using the standart install (linux reiserfs).
> 
> We do not make changes on kernel or setup.
> 
> Is it good idea to change redhat?
> 
> Also it is possible to use redhat AS 3.0, but we do not find any info how to 
> instsall it with reiserfs. Exist any problems on this distro?
> 
> regards,
> ivan.
-- 
Regards, Redeeman
()  ascii ribbon campaign - against html e-mail 
/\                        - against microsoft attachments

Re: production use of reiserfs

[Chris Mason]

On Mon, 2003-12-01 at 08:33, Chris Mason wrote:

> Hans, please send along details for the xattr bugs you think are in the
> suse kernel.  We have no open bug reports for them, and they work pretty
> well.

That's what I get for sending email before reading my entire inbox.  I
found the recent XATTR thread, I thought we had resolved that but I'll
wait for Jeff to comment.

-chris

Re: production use of reiserfs

[Bennett Todd]

[-- Attachment #1: Type: text/plain, Size: 1291 bytes --]

2003-12-01T10:19:39 Redeeman:
> i feel that the linux filesystems is much more unstable after a
> power failure than windows filesystens, but in advance they are
> much faster.

In advance they're much faster because they're much more
sophisticated, with a great deal of design and implementation
focused on making them fast. Until this additional complexity is
very well debugged, they can be more fragile. ext2 is quite robust,
and I've found ext3 to be bulletproof on crashing machines.

Reiserfs is more sophisticated yet, but it has hit a quite stable
plateau; current kernel + current reiserfsprogs = very robust, with
one notable caveat: reiserfs does more critical and elaborate
computation than other filesystems, so it has acquired a reputation
for flushing hardware bugs out of the wainscotting.

If you're using reiserfs in a current prod kernel on a filesystem
built with current reiserfsprogs, and you see a crash, look for a
hardware problem. If reiserfs crashes when ext3 doesn't, look
especially hard for marginal CPU or memory that gets flaky when it's
rode hard.

> but ofcourse, windows filesystems are made to be crash proof.

If Linux crashed as often as Windows, all our filesystems would be
as robust in the face of crashes as vfat. Practice makes perfect.

-Bennett

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]